Occasional Paper 14 - DECENTRALIZATION IN CHILE

DECENTRALIZATION IN CHILE

Frances Stewart and Gustav Ranis, 1994


  1. Introduction
  2. Decentralization and the structure of Government: An Overview
  3. Specific structure of decentralization in Chile
  4. The Impact
  5. Conclusions
  6. References
  7. Tables


I. Introduction

Major changes in the political regime and decentralization strategy over the last few decades make Chile an interesting case-study of decentralization. Chile is a "narrow" nation covering almost three-hundred thousand square miles, ranging from coastal lands to the Andean mountain range. Its population is fairly homogeneous, concentrated in the central regions and metropolitan areas, with no substantial ethnic or cultural differences. The government has been unitary since political independence in 1810. There exists considerable variation in economic resources and production processes across regions, with agriculture, fishing, forestry, and mining the main industries.

Chile's social and economic programs have a solid foundation that dates back to 1900, including long-term projects in housing and water supply. These programs were national in origin and quite independent of local authorities. Over the last three decades Chile has undergone radical changes in the structure of its government, including in the approach to decentralization. However, capturing the full effects of the changing scene in Chile is particularly difficult due to the fact that the most recent political changes occurred as recently as 1990 and some of the structures of the new democratic government are still being debated in the parliament. Throughout the various past regimes, Chilean national policy-makers have put a high value on health and education and have attained excellent results. For example, Chile's Human Development Index in 1990 stood eighth among developing countries (and first in Latin America), with a life expectancy of 72, an under five mortality rate of 25%, and adult literacy at 93% (Table 1 summarizes).

Chile has experienced major changes in political regime in its recent history. Frei led the Christian Democratic government (with political roots dating back to the 19th century) until 1970, when he was voted out and replaced by Allende, a socialist. Allende was deposed by Pinochet in a military coup in 1973, and Pinochet ruled Chile under a highly repressive military regime until 1989. Democracy returned with Aylwin, a Christian Democrat, leading a coalition of center and left parties. Pinochet, however, retained control over the Army and some of the nation's policies via the considerable veto power of his parliamentary appointees.

These dramatic political changes have been paralleled by changes in economic policy. The Frei regime followed an import-substituting industrialization policy, typical of Latin America at that time. That policy was accompanied by an expansionary monetary policy and a fairly high (but still tolerable) level of inflation. Allende pursued more populist policies, including large wage increases, which led to a high rate of inflation. That, and the attempted nationalization of foreign-owned copper companies, were the main factors that led to the military putsch by Pinochet. In a somewhat wavering way, Pinochet adopted monetarist, pro-market and open-economy policies favored by the "Chicago boys." The strict monetarism, accompanied by fixed exchange rates, proved stagnationist for many years, with growth of less than 1% throughout the 1970s and negative in the early 1980s. Inflation was reduced, but unemployment exploded (up to 17% in 1985). The poverty rate also rose sharply, from below 20% in 1970 to nearly 40% for much of the 1980s, with many of these problems exacerbated by fluctuating trade policies and downturns in the copper market. Beginning in 1986, however, there was a high rate of both output growth (4.4%) and natural resource-based export growth. Unemployment rates dropped dramatically (to 6%) and there are signs of a fall in poverty (Table 2). The shrinking of the economy in the early 1980s also led to an increase in the variance of regional per capita income and welfare that has not yet abated.1

The various regimes also made changes in social policy that affected the nature and impact of decentralization in Chile. Under Frei and Allende, there existed a welfare state type of approach to social policy, with the objective of free and universal entitlement to state-provided health care and education. Pinochet introduced a radical change with emphasis on the private provision of health and education services, targeting state assistance to municipalities which fell below the poverty line. For primary and secondary education, a subsidy voucher system of government support was introduced on a per pupil basis, paid directly to the schools, whether public or in the private-subsidized sector, while state support for tertiary education was sharply reduced. In health, people were encouraged to opt out of the public system by paying contributions into private insurance groups. About 20% of the population chose to do this, especially among the upper income groups, removing nearly 50% of the financial burden from the public health system. During the 1980s, health charges were introduced in the public health system, with exemptions for poorer groups, mothers and young children. A complex program of inquiry into family circumstances, the questionnaires of the CAS (Comites de Asistencia Social) administered by the municipalities, permitted fairly efficient targeting of social assistance. This system, together with emergency employment schemes, prevented total destitution during periods of generally rising poverty and unemployment.

In what follows we review the changes in decentralization that have accompanied the changes in political regimes and some of their effects on various dimensions affecting the human condition. Section II provides an overview of the decentralization process, while Section III provides a more detailed discussion of the nature and extent of decentralization achieved, both under Pinochet and since, including changes in the system currently being instituted. Section IV assesses the qualitative and quantitative impact of these decentralization measures; Section V presents some conclusions.

II. Decentralization and the structure of Government: An Overview

The nature of Chilean decentralization over time was affected by changes in the political environment. During the democratic governments of Frei and Allende there were elected municipal councils and various neighborhood groups. These groups served as the base of popular participation in national affairs and often were organized around local and regional issues. Allende, for example, used local groups to advance social programs, especially the neighborhood food distribution program. While democratic structures were in existence at the municipal level under both Frei and Allende, the resource base of the municipalities was weak.

Pinochet abolished all such democratic structures and replaced them with a military hierarchy in which centrally appointed intendentes were responsible for regions, governors headed provinces, and alcaldes (mayors) were responsible for municipalities. All appointees were named directly by the President and were loyal to the national government first, their area second. This effectively brought all municipalities under direct national control. All the intendentes and governors (and most of the mayors) came out of the military, and the clear aim of the new structure was to transmit commands from above most efficiently. 2 The provision of public services in Chile under Pinochet was intended to operate like a private market, with strong central control. Municipal-level tasks were privatized wherever possible and the streamlining of all levels of government became a primary goal. Municipal governments thus acted like 'service delivery agents,' providing local public service on a cost-effective basis, without having local governing power. This form of decentralization created a direct line of hierarchical control from Pinochet down throughout the nation, consistent with his political and macroeconomic agendas.

Pinochet gave this new structure of government greater powers and resources, especially through the 1980s. The most important changes were the delegation of responsibility for primary and secondary education and primary health care to the municipalities, for which funds were transferred on a per capita basis. In addition, extra resources were provided to the municipalities through the property tax (via the Common Municipal Fund (FCM)), additional license income, and the transfer of investment resources through the Regional Development Fund (FNDR). 3 The resources owned by the municipalities thus increased significantly over the 1980s, although they remained a very small proportion of total government revenue (around 6%). The municipalities were, moreover, subject to strict central government guidelines and restrictions in terms of finance, were not permitted to run deficits, to have discretionary local tax control, or to take a significant role in local economic development. There were some redistributive features in the system of finance for both the municipalities and the regions. Among the municipalities this was achieved by the operation of the Fondo Comun Municipal and the Regional Development Fund. The Pinochet government also made considerable efforts to professionalize the staff of municipalities via increased training and objective hiring standards.

The major change in decentralization achieved by the current democratic regime has been to institute elections for municipal councils and mayors (the first elections occurring in 1992) and to give the municipalities additional finance. But the extent to which municipalities are free to raise additional funds is quite limited, and tax base inequalities raise distributional problems. There are additional proposals under consideration to strengthen and democratize the regional level of government. While the intendentes will continue to be appointed by the President, new regional councils will be elected by the municipal councils.

The provincial level of government has been largely an irrelevance over the years. It has been retained in the recent reforms with an appointed Governor and an advisory council. However, it has neither significant powers nor any source of finance. Its influence is limited by the personality of its Governor and the proposals it makes to the regional government. It occasionally assists the regional government in administrative tasks or (rarely) coordinates municipal actions. Because of its limited influence, we ignore it throughout what follows.

One observer summarized the changing scene at the local level as follows:
Frei I = democracy without modernization
Piochet = modernization without democracy
Aylwin and Frei II = modernization and democracy

III. Specific structure of decentralization in Chile

During the Pinochet rule, strict military-style control and an ideological desire to reduce the size of the state (through both decentralization and privatization) were the major motives for the somewhat odd combination of central control at the national level and decentralizing structures through the municipalities. At the same time, there were efforts by foreign-aid donors and NGOs to bypass the Pinochet structures, leading to a variety of specific funds. More recently, the democratization of government has led to a debate on the size, shape and resources of different levels of government which has focused on (a) the desirable level of political control at different levels; (b) the desire to provide a firm foundation for sustained democracy; and (c) the desirability of increased decentralization from the point of view of efficiency, equity, participation and accountability.

In the more detailed description of the structure of government contained in this section, we highlight important differences between the Pinochet structure and the post-1990 structure. The section is organized as follows: first, the administrative structure is discussed; second, the financial structure is analyzed; third, decentralization ratios are reviewed; fourth, we study changing human capacities at the local level; finally, we conclude with an assessment of the current state of decentralization and how proposed changes are likely to impact upon it.

A. Administrative structures.

The Pinochet regime established four tiers of government: the central government; 13 regions (including the Santiago Metropolitan Region -- SMR), 51 provinces (with only nominal functions) and 325 municipalities. 4 The structure is summarized in Figure 1. This basic structure has been maintained, but with some changes, by the new democratic regime (Figure 2).

A.1. Regions.

The regions vary in size, population and income (Table 3). Population density differs from 0.7 persons per Km2 (XI) to 341 in the SMR; total population per region from 80,000 (XI) to 5.2 million (SMR); income per capita from $791 (IX) to $3,431 (XII); and the number of people falling below a minimum consumption standard from 1.9 million (SMR) to 18,000 (XI).

Each region is headed by an appointed intendente. Under Pinochet this was a military appointment, not necessarily made from within the region. In the Pinochet structure (Figure 1), the intendentes were advised by Regional Development Councils (COREDEs), which consisted of appointed representatives of local groups (including entrepreneurs and workers). Local representatives of the sectoral ministries (SEREMIs) helped formulate and execute decisions, while employees of the planning minister at the regional level (SERPLACs) evaluated project proposals.

The Pinochet-appointed intendentes were replaced by Aylwin by civilians from within the regions (Figure 2). In 1993, the advisory councils were replaced by Regional Councils whose members are elected by municipal councillors, with responsibility for approving regional development plans and allocating the resources assigned for regional investment. Appointed social and economic advisory councils have been instituted at the provincial level, and the remaining structure of SEREMIs and SERPLACs at the regional level are unchanged. It should also be noted that all technical officials at the regional level remain employees of the central government and not of the regional governments.

During Pinochet's reign, the main functions of the regional level of government were to consider regional investment decisions, especially the distribution of the FNDR, and to provide an efficient central mechanism to oversee the decisions of the municipal mayors. In theory, the new regional governments have much the same responsibility although their financial responsibilities have been extended to the programmes of Sectoral Regional Investment (ISAR).. In practice, differences will be manifested only if the regional government increases its control over resources. However, the "chain of command" aspect of the military regime -- from the President through the intendentes to the mayors -- has been weakened. The election of mayors responsible to the local councils and the electorate, not under the command of the President, has increased the probability of regional governments being able to take a significant role in fiscal matters.

A.2 Municipalities.

The municipalities, like the regions, display huge variance in size and socio-economic characteristics. Forty out of the 3345 have more than 100,000 inhabitants, while 200 have less than 20,000. The poor are disproportionately concentrated in the largest municipalities6 (Table 4).

Although the municipalities were governed undemocratically under Pinochet, they had some autonomy in the sense that they were entitled to certain tax revenues which they could dispose of more-or-less as they pleased, subject to general guidelines. However, their officials, both with respect to particular functions and for project evaluation (the SECPLACs), were appointed by (and responsible to) the national government. The Pinochet regime provided some additional funds to the municipalities and delegated responsibility for primary health care and basic (primary and secondary) education to them. The municipalities were also made responsible for the poverty enquiry (CAS) and for administering a collection of poverty-relief programs, including the emergency employment program. Their other functions included the paving of urban and rural feeder roads, solid waste collection and disposal, public transportation, drainage, street lighting, parks and recreation, and public cemeteries.

The post-1989 government has democratized the municipalities, providing for elected councils and mayors, 7 but retaining advisory social and economic councils consisting of representatives of interest groups. The basic functions of the municipalities remain unchanged, but their funds have been increased by a small amount and they have been given a limited amount of increased discretion over local tax rates.

B. Financial structure.

As noted above, the regional governments have no funds of their own, nor a budget, although they (especially through the SERPLACs) play a part in allocating the regional investment funds. The focus here, therefore, must be on the municipality.

Historically, the municipalities have long had their own source of funds, varying from licenses to the sale of goods and services. This funding base brought in extremely little revenue and was regressively distributed, with wealthier municipalities having much more per capita resources than poorer ones. In 1977-79, the municipal share of total government revenues was just 2.5%. The Pinochet regime provided new revenue sources and changed the distribution of revenue among municipalities. Municipal governments, however, remained highly dependent on central government revenues. Some of Pinochet's innovations included:

i) From 1979 all revenue from the property tax was allocated to the municipalities (which had previously received only a fraction). To a certain extent this displaced the "aporte fiscal," a central government transfer to the municipalities which was sharply reduced in value, then phased out completely by 1986. This additional revenue from the property-tax-cum-FCM was much greater than the aporte fiscal and accounted for 34% of revenue from the municipalities' own taxes and licenses in 1990, the biggest single source of revenue. This tax, however, was not under local control as both the tax rate and assessments were set by the central government.

ii) In the previous system, both "own" revenue and finance from the central government varied with the wealth of the municipalities. Thus in 1974, the most wealthy 10% of the municipalities received 70% of the total income. 8 A system of redistribution among municipalities was introduced through the FCM. It was financed by 60% of the property tax, 50% of the vehicle tax in all municipalities, and a proportion of receipts from a tax on business in three high income municipalities. The FCM was distributed in accordance with a formula, including a flat rate per municipality, the number of inhabitants per municipality, the number of exemptions from the property tax (a poverty indicator since a flat rate value of property determines exemption), and a proportion according to the shortfall of municipality income from average municipality income. The remaining 10% of the FCM was conditionally allocated to municipalities with emergency deficits. This fund has been an effective means of redistributing revenue from communities with a strong tax base and targeting those with a weak tax base.

iii) Funds to finance basic education, health and social assistance were transferred with the transfer of responsibility for these services during the 1980s. The funds for social assistance exactly matched the expenditures. In education, a flat rate subsidy per pupil was paid, but the real value of these subsidies fell quite sharply over the 1980s. 9 Health payments were made in accordance with specified services provided. Major municipal deficits emerged due to the reduced real value of the education subsidy and the accompanying privatization which removed pupils (and their subsidies) from the municipal schools while costs did not fall proportionately. In health, the main problem was a ceiling imposed on total expenditures in 1983, while health services expanded and the real value of government finance again fell. 10 The 'flat' payments to cover health and education expenditures were meant to provide local government with an incentive to control costs. However, these reimbursements did not, in the aggregate, cover actual costs. This caused an acute financial squeeze on the municipalities, who had to use some of their "own" funds reallocated from other investment and operating areas to finance health and education. Some municipalities also ran deficits, although in principle these were prohibited. Ways of financing deficits included using a portion of the FCM reserved for emergencies and petitioning for special funds established by the Pinochet Government. Since 1990, there have been occasional overtures to the Finance Ministry, which has funded deficits in a few cases but has simultaneously applied "IMF-conditionality" to the municipalities to discourage repetitions.

iv) The Pinochet regime also initiated a flow of funds for investment, the most important being the Regional Development Fund (established in 1975) amounting to 15% of the public sector investment budget (largely financed by the IDB). This fund finances new investment projects in health, education and minor infrastructure, such as rural roads, and is distributed in response to project submissions by the municipalities and regions. The projects are subject to an appraisal by the planning ministry. Small projects can be evaluated and approved at the regional level, but projects above about U.S. $200,000 must be analyzed by the center. Redistributive criteria, including the population size and the inverse of per capita income of the area, in principle determine the regional distribution of the funds, but they are not always applied in practice. 90% of the Fund is allocated to five areas: basic education, basic health, sanitation, rural roads, and small urban roads. The remainder is discretionary and can be spent on a wide range of projects.

v) Other investment funds for use by municipalities -- with varying degrees of local control -- were set aside for neighborhood improvement, urban improvement, communal equipment, sanitation, pavements, and sports. Each program is designed to cover a particular type of activity or need. The Regional Fund is probably most responsive to municipal submissions, although both regional and central government play a critical role (Figure 3). For some of the funds, the role of the municipality is quite marginal: for example, in the neighborhood improvement fund, the role of the municipality is confined to identifying poor areas, with priorities set at the regional level. 11 Municipalities typically are only responsible for execution of the projects.

Proposed changes in the financing system include a revaluation of the property tax (including an increase in the exemption level), discretion to the municipalities to alter the rate of the property tax between 1.1% to 1.6% (with municipalities charging less than 1.6% having their FCM reduced proportionately), removal of a ceiling on the business tax, and small changes in the formula for distribution of the FCM (the most important being the inclusion of the percentage of uneducated mothers and child malnutrition as criteria). It is expected that these changes will increase municipal own revenue by 20-30%, giving more to larger/more developed municipalities. 12

A new scheme initiated at the regional level, known as ISAR, allows the transfer of a small proportion (up to 5%) of the investments of sectoral ministries to the regional governments at their request. Programmes transferred under the ISAR mechanism include rural and urban roads and pavements, neighbourhood schemes and potable water as well as various projects of the Social Investment Fund (FOSIS). The Regional Fund (FNDR) is to increase by 25% above the increase in total public sector investment each year over four years, with some of the additional funds to be spent by the regional governments rather than the municipalities. Thus the proposals give more investment resources to the region, but none of them are totally under regional control; the regional governments still lack any source of current expenditure.

As can be seen from Tables 6 and 7, the municipalities are heavily dependent on outside money, especially for investment finance. "Own" revenue, including taxes and FCM, accounts for 57% of the total budget revenue of the municipalities. But the municipalities only finance an estimated 16% of the investment in their area, an amount which just exceeds the contribution of the various external funds. Both are dwarfed by the expenditures of the central ministries, which account for nearly 70% of the investment in the regions. One future consideration which may alter the central government/municipal fiscal transfer relationship concerns the FCM. Its redistributive nature was 'set in stone' under the strict rule of Pinochet, but this could be subject to resistance with the increased democratization of local government.

Overall, Chile faces a large local revenue dilemma, as per-capita municipal revenues vary by a factor of more than 100 to 1 among the cities alone (the variance is even higher when rural municipalities are included). This implies that merely assigning discretionary tax control to municipal governments will not alleviate the built-up fiscal pressures and will only increase the variance in municipal welfare. Correcting the inequalities inherent in local tax bases -- by some system of transfer from the central government -- must be a major part of any decentralization effort granting more power to the municipal governments.

C. Decentralization Ratios.

Revenue and expenditure decentralization ratios were very low in Chile before the reforms of the 1980s: only 2.3% of total government revenue was raised by municipal taxes and the municipalities were responsible for an even smaller proportion of total expenditure (less than 2%, see Table 8). The financial autonomy ratio was high, however, at 0.92, indicating the small amount of government transfers to the municipalities. Revenue and expenditure ratios increased sharply during the 1980s: municipal revenue rose to 6% of total government revenue by 1988, and if health and education are included, the expenditure ratio rose to almost 8.3% The financial autonomy ratio fell sharply since health and education are financed by transfers from the central government. In 1990, the municipalities financed 60% of their activities from their own resources. However, there was some decline in the decentralization ratios between 1988 and 1990.

The transfer of responsibilities for education and health to the municipalities led to a sharp increase in the sectoral decentralization ratios, especially in education where, by 1988, the municipalities were responsible for one quarter of government expenditure (Table 9) falling back to around one fifth by 1990. In health, the proportion was only 3.2% in 1990, reflecting the low proportion of health resources devoted to primary health care, the responsibility of the municipalities. The infrastructure decentralization ratio rose from just over 2% in 1977 to nearly 25% in 1988 and 39% in 1990, reflecting the increase in municipalities' investments accompanying their increase in financial resources.

D. Human Capacities at the Local Level.

A major thrust of Pinochet policies was to "professionalize" the municipalities, increasing their qualified personnel and reducing the numbers of unqualified employees. The expected efficiency gains from a smaller, but better trained and paid, staff led to this decision, as local governments were considered overstaffed and poorly managed. The reductions occurred by national edict via the 1976 Municipalities Act. This change in personnel composition became especially important as new functions were transferred to the municipalities and as access to resources from the special funds became dependent upon the presentation of good projects to the national funding agencies. The career structure of municipal employees was improved through increased training and tenure. Part of the professionalization consisted of heavy emphasis on cost-benefit analysis by SECPLAC and SERPLAC personnel, with training provided by the central government. Other types of training were provided by the municipalities, with support from external agencies.

Municipalities were encouraged to restrict total employment by a rule which put a maximum of 35% on personnel costs in the municipal budget (not including health and education) and a ceiling on the ratio of municipal employees to population. One of the biggest changes brought about by Pinochet at the municipal level was the greatly reduced role of the local government as an employer of unskilled labor (excluding the emergency employment schemes which temporarily re-employed many of those who lost their municipal jobs). The overall aim was to encourage the contracting out of services to the private sector.

Aggregate figures suggest that the increased professionalization of local government employees was effective. Professional and technical staff rose from 8% to 33% from 1975 to 1988, a rise almost exactly offset by the fall in auxiliary (untrained) employees (Table 10). Overall, municipal employment fell 15% from 2.01 per thousand in 1975 to 1.72 per thousand in 1988.

While the professionalization of the municipal personnel occurred across the board, there is still a large discrepancy in human capacities among municipalities, with smaller/poorer municipalities much weaker than larger/richer ones. Of the total, 118 municipalities are estimated to have "low" economic management capacity, 134 "medium" and 73 "high".13 The worst concentration of low capacity is found in Regions IV, VII and VIII (over 60% with low capacity). These regions all have below average per capita incomes and above average poverty rates (Table 3). Exceptionally good capacity is found in Regions I and III (both regions have above average per capita incomes, but III has a high poverty rate).

E. Assessment of the Current Extent of Decentralization.

In making this assessment, we adopt the accepted distinction between deconcentration (central government posts employees at the local level), delegation (central government delegates its powers to representatives at the local level), and devolution (the central government completely cedes powers to local decision-makers). 14

There was minimal decentralization during the Frei/Allende administrations, shown by the few functions and very limited resources available to the municipalities. The small amount there was can be described as constrained devolution. The municipalities were autonomous units with elected governments, responsible to the local electorates and not to the central government. However, they were heavily constrained by limitations on their tax powers and functions.

The Pinochet period showed a major increase in decentralization, as indicated by the new functions and resources assigned to the municipalities and regions. This decentralization was strictly deconcentration and delegation, as abolishing the elected authorities at municipal levels eliminated all elements of devolution. Officials at every level were appointed by (and responsible to) the President. Constraints were imposed on the use of resources (e.g. employment restrictions), and the new investment funds involved strong central influence. A variety of central rules was imposed on local government decisions regarding education and health. For example, the curriculum was determined centrally, textbooks were produced centrally, and health services were required to impose user charges.

The democratic government has reinstituted elements of devolution at the municipal level with the general election of mayors and councils. Moreover, it has sustained the resource improvements of the Pinochet era, and present proposals will further increase both own-funds of the municipality by 20 to 30% and the size of the Regional Fund. Thus, current decisions appear to maintain and even increase the decentralization of the 1980s, while transforming the previous deconcentration/delegation into devolution. The devolution remains constrained, however, as the municipalities have been permitted to vary taxes by only a very small amount and with penalties if the property tax rate falls below or above a certain rate. Resource access is still limited by central government decisions. There is no change in the operation of the Regional Fund to give more control to the municipalities, and the municipalities may lose resources to the regional governments. Finally, no significant relaxation of central rules on the municipal use of own-funds or in the education and health sectors is being contemplated.

The democratic government proposes to strengthen the regional level of government by providing resources through the Regional Fund and ISAR, and by democratizing (albeit indirectly) the Regional Councils. Both sources of funds are subject to heavy central control, as all large projects for the Regional Fund go to the central level while ISAR funds are allocated to specific areas. The regional governments will have annual budgets but no independent source of funds, only annual central government transfers.

The intendentes for now remain Presidential appointments, with the difficult dual role of representing the President in the region and heading the regional government. The intendente's "cabinet" consists of regional representatives of the sectoral ministries and a project evaluation staff which will be based in the Planning Ministry. The regional governments have no actual employees of their own. The powers of the Regional Councils are confined to rejecting/reducing (not increasing) the regional budget. The regional government proposed thus represents constrained delegation, with limited financial powers. However constrained, it can be interpreted as an increase in decentralization of the delegation kind. If the changes take resources and/or functions away from the municipal level, they are a backward step from the perspective of decentralization. This is likely to be true to some extent of the use of the enlarged Regional Fund. In contrast, if the new resources are at the expense of central government ministries, (i.e. ISAR), they increase the extent of decentralization.

In summary, the pre-1974 situation of very limited devolution was replaced by increased decentralization of a deconcentration/delegation type. Constrained devolution has recently been reintroduced at a much more significant level in terms of resources and functions than pre-1974. However, changes at the regional level remain ambiguous in their effect; they involve increases in delegation but may do so at the expense of the powers and resources of the much more autonomous municipalities.

 

1 In 1980 the annual standard deviation in per capita regional income was 21,618 Chilean pesos (1977 $), while by 1986 this figure rose to 22,001 Chilean pesos (1977 $) without a corresponding rise in the mean. Bulnes (1988), World Bank (1992b).

2 According to a World Bank report on the Pinochet structures: "the particular philosophy advanced by Chile's military regime [is] one in which local government is a decentralized instrument for implementing central government policy." World Bank, 1990b, Ch. III, p. 1.

3 Fondo Nacional de Desarrollo Regional

4 The innovation here was the regional level; the previous structure had included provinces and municipalities.

5 There were originally 325 municipalities, but nine new ones were recently created in the metropolitan region of Santiago.

6 World Bank, 1992.

7 Any councillor who receives more than 35% of the vote automatically becomes mayor. The remaining mayors are to be elected (from their number) by the elected councilors. The system thus contains elements of direct election (as in the US) and indirect election (as in the U.K.).

8 Irarrazaval, 1992, p. 104.

9 Public expenditure on basic education fell by a quarter from 1981-89, with a similar fall in the education subsidy going to the municipalities. (Espinola, 1991, p. 26, 97).

10 It is estimated that the real value of the general tariff schedule for health fell by 60% between 1983 and 1990. (Irarrazaval, 1992, p. 174).

11 ILPES, 1992, p. 190.

12 World Bank, 1992, p. 38.

13 The index is made up by weighting own incomes, dependency on the FCM, municipal and total investment. (Duran et. al., 1992 p. 25). It is therefore an indication of the economic capacity of the municipalities, rather than their human management capacity.

14 Rondinelli et al., 1983.